How Oscar's Grind Works in Roulette

Oscar's Grind is a betting progression targeting a one-unit profit per cycle on roulette even-money bets (1:1 profit odds); it changes stake sizes and their order, but cannot change outcomes or the house edge. After a full loss, the stake stays the same. After a win, it increases only as far as the amount needed to reach the cycle's +1-unit target.

This is a stake-setting rule, not a way to predict spins or guarantee profit. The classic rules below assume that zero causes a full loss on the selected wager.

How to run a classic cycle

Choose a base unit, start the cycle with a net result of 0, and stake 1 unit. After each spin, update the cycle net result before calculating the next stake.

Your base unit is the reference amount for your stakes. If 1 unit is $10, a 2-unit stake is $20. It is an accounting choice, not a bankroll recommendation.

The cycle net result is the cumulative profit or loss within the current cycle. It is separate from the next stake. At 1:1 profit odds, a win adds the stake to the cycle net result; a full loss subtracts it.

Use an even-money roulette bet such as red or black, odd or even, or 1–18 or 19–36. Check the current table rules, particularly how zero is settled.

The exact rule after a win in an unfinished cycle is:

Next stake = min(previous stake + 1 unit, 1 unit − current cycle net result).

The second term caps the stake at the amount needed to finish at exactly +1 unit on the next win. For example, when the cycle net result is 0, the next stake is 1 unit—even if increasing the previous stake would otherwise produce a 3-unit stake.

These mechanics are described in the Wizard of Odds rules for Oscar's Grind. They assume each calculated stake can be placed; personal limits and table restrictions may interrupt the cycle.

Worked example: L, L, L, W, W, W

For three losses followed by three wins, the stakes are 1, 1, 1, 1, 2, 1 units, and the cycle finishes at +1 unit.

Assume a base unit of $10, full losses on losing spins, and 1:1 profit odds on wins. All amounts in this example are exact; no rounding is needed.

After spin 4, increasing the stake from 1 to 2 units is within the cap: the cycle still needs 3 units to reach its target. After spin 5, the cycle net result is 0, so the cap makes the next stake 1 unit rather than 3.

This particular cycle has:

• Turnover: 7 units, or $70.
• Maximum stake placed: 2 units, or $20.
• Cycle profit: 1 unit, or $10.

Turnover is the sum of all stakes placed—not profit, the next stake or the bankroll required at the start. This is a mechanics example, not a forecast of cycle completion, average duration or future results.

Why 1:1 profit odds do not remove the house edge

An even-money bet pays 1:1 profit odds, but it does not have a 50% win probability: with full losses on zero, there are more losing outcomes than winning outcomes.

The following models distinguish expected value (EV) per 1-unit stake from house edge, the mathematical expected loss rate on turnover.

La Partage applies only when the current table rules explicitly grant it on the selected even-money wager. Under that rule, zero loses half the stake. It is not a universal feature of single-zero roulette.

The official Loto-Québec roulette rules provide a table-rule reference, but the document is jurisdiction- and version-specific. Check the current rules of the table you intend to use.

The classic whole-unit progression above does not define a transition after a half loss. Applying La Partage therefore requires a separately defined accounting and next-stake rule; do not silently treat a half loss as a full loss.

How turnover relates to expected loss

For specified rules, expected loss is calculated as turnover × house edge. It is a mathematical expectation, not a prediction of the realised session result.

For 100 fixed spins at $10 each, turnover is $1,000. Using the exact house-edge fractions and rounding the resulting amounts to the nearest cent gives:

These calculations use a fixed stake and a fixed number of spins. They are not calculations of an Oscar's Grind session with changing stakes. Expected loss is shown as a positive amount; the corresponding EV of the net result is negative.

House edge is not the realised session result. Finishing an individual cycle in profit does not alter the game's mathematical conditions. The house edge guide explains that distinction further.

How to use the House Edge Calculator

The existing House Edge Calculator can estimate expected loss for a fixed stake and number of rounds, but it cannot check the Oscar's Grind progression.

1. Open the House Edge Calculator.
2. Select European Roulette with the 2.70% house-edge preset.
3. Set the stake per round to $10.
4. Switch the session duration to rounds.
5. Enter 100 rounds.
6. Read the result: $1,000 total wagered and −$27.00 displayed as expected loss.

Local House Edge Calculator inputs for European Roulette, 10 dollars per round and 100 rounds

Local interface, not production: the 2.70% preset and $10 across 100 rounds give $1,000 turnover and a displayed result of −$27.00; the exact 1/37 model gives −$27.03, rounded to cents.

Local House Edge Calculator results showing 1000 dollars total wagered and minus 27 dollars expected loss

Local interface, not production: $1,000 turnover produces −$27.00 with the 2.70% preset; the exact 1/37 model gives −$27.03, rounded to cents.

The difference comes from input precision: the calculator uses a rounded edge of 2.70%, while the exact single-zero model uses 1/37. Both figures describe expected net loss, not what a particular session will lose.

The calculator uses fixed stake × rounds. It does not track the cycle net result, calculate the next stake or validate Oscar's Grind transitions.

For a tracked session with variable stakes, first sum all actual stakes to find turnover. Use the calculator only if that turnover—or the actual average stake multiplied by the number of rounds—can be represented by its inputs. Even then, it estimates expected loss rather than checking the progression.

How Oscar's Grind differs from flat betting and Martingale

The mechanical difference is when stakes change: flat betting keeps them constant, classic Martingale doubles them after losses, and Oscar's Grind holds them after full losses and raises them after wins, subject to its cycle cap.

This compares mechanics only—not safety or profitability. If a cycle starts with a 1-unit opening stake and the first ten bets are consecutive full losses, the next stake is 1 unit under Oscar's Grind and flat betting, and 1024 units under Martingale doubling. The chart compares only the next stake after this specific sequence under the named rule variants, not the maximum stake possible in arbitrary play. It does not show winning probabilities or rank the systems by risk, safety, expected value, or profitability.

The model assumes independent spins: past outcomes do not predict the next outcome, and a losing streak does not make a win due. Changing the stake does not change the probability of the selected outcome.

What to do about limits and unfinished cycles

Set a session stop rule before playing and stop when it is reached, even if the cycle net result remains below +1 unit.

Before the first stake:

• Choose an affordable base unit.
• Check the table's minimum stake, maximum stake and permitted increments.
• Set money, time and maximum-stake limits.
• Commit to stopping at those limits rather than extending them to finish a cycle.

If the next calculated stake exceeds a personal or table limit, the cycle may remain unfinished. Do not chase the +1-unit target or replace the stop rule with an off-system increase.

There is no universal bankroll multiple, “safe” bankroll size, guaranteed maximum cycle length or established average cycle duration for every session. A session limit controls participation; it does not change the house edge or guarantee profit.